Change detection of buildings in urban environment from high spatial resolution satellite images using existing cartographic data and prior knowledge

Several studies in remote sensing image processing have tackled the issue of change detection for cartographic needs. Many algorithms have been developed, but very few have been applied for urban studies to update maps from high-spatial-resolution remote sensing images. The semantic richness of the image increases and makes image analysis more difficult. Change detection from high spatial resolution images such as Ikonos and QuickBird is even more challenging, especially in complex environments like urban areas characterized by small objects such as houses, individual trees and roads, and by shadows. In addition, even if they are usually available, existing digital map data often are not incorporated in all steps of change detection process. This research project proposes a new object oriented method for the detection of building changes from high-spatial-resolution images in urban areas inspired by the specific problem of guiding the process by using existing cartographic data and knowledge. The use of the existing digital map and the integration of existing knowledge allow optimizing the change detection process on the image while offering the possibility to target and to accelerate the research of the changes.